A founding year invites a tidy story: "in business since 2004, so twenty-odd years of data". A company that has changed its core system three times has a less tidy one. The current system may hold only what each migration chose to carry across, and the rest may sit in an old tenant, a folder of exports or nowhere at all.
Count the cutovers, not the years
Each cutover is a filter. A migration project decides which history moves into the new system and in what shape, and that decision is seldom written into anything a portfolio operations analyst would see.
Microsoft's documentation for moving from Dynamics GP to Business Central online shows how specific that choice can be. Detailed historical transactions come across as a separate "GP historical snapshot", held in extension tables and shown on their own list pages. The company sets an Oldest Snapshot Year that chooses the earliest year whose history is brought over, and the snapshot runs as a background job after the main migration finishes. How deep the new system's history goes can come down to one field someone filled in on a migration screen.
Other products handle history differently, which is the point: the analyst asks, and does not infer depth from a vendor name or a founding year.
The minimum context for each era
Collect five facts per era through whoever holds the CEO relationship, in the company's words:
- Years, start and end, each marked reported or estimated.
- System, by type and plain name, and the record types it held (orders, tickets, projects, claims).
- Administrator then, and who can speak to it now. For older eras these are often different people.
- The owner of the company at the time, if it has changed hands; rights can follow ownership.
- Where the era's records are today, in one phrase.
Add the current full-time headcount at peak (contractors excluded) and the CEO as sponsor, with whether the CEO wants a conversation. That is enough to draw the canvas.
The system-history timeline canvas
Draw one horizontal line. Put a box on it for each era and a gate between boxes for each cutover. Under every gate, write the answers to five crossing questions; under every era, write what stayed behind.
| Crossing question | Answers you will hear | What it means for the summary |
|---|---|---|
| 1. What crossed? | Full history; open items and balances; a snapshot from a chosen year; nothing (a fresh start) | Full history and snapshots carry records forward. Balances carry totals, not the work behind them |
| 2. Did the IDs cross? | Same IDs; new IDs with a mapping table; new IDs and no mapping | Without a mapping, the same customer or order in two eras cannot be joined, and the history splits |
| 3. What stayed behind, and in what form? | Old system running read-only; export files; a database backup; a vendor tenant until contract end; nothing | Each form has its own keeper and its own end date |
| 4. Who can open it today? | A named person, and what they need: a license, a server, the right software version | A backup nobody can restore is not history anyone can describe |
| 5. When does access end? | A contract end date; a planned switch-off; disposal of retired hardware | A dated end is real timing, and belongs on the canvas |
Question 5 has a hard floor. NIST defines sanitization as a process that renders access to target data on the media infeasible for a given level of effort. Once a retired server's disks have been sanitized, that era survives only in whatever crossed a gate or was exported before the wipe. Asking "were the old servers wiped, and when?" is not pessimism; it dates the end of an era.
Method note. The canvas and its five questions were built for this page from the published screening criteria, the program terms, Microsoft's migration documentation as one example of how history crosses a cutover, and NIST's definition of sanitization, all checked 2026-10-27. It records where each era's records are and who can reach them. It does not measure demand, record quality, rights or payout, and it does not judge whether separate eras are useful together; SourceX assesses that in review.
One distributor, four eras (Illustrative)
Illustrative. The distributor, its systems and its dates are invented for this example.
A 260-person industrial distributor, founded in 2004, has changed its core business system three times. Its controller and IT lead filled in the canvas from memory and old project files.
| Era | Years | System | At the next cutover | IDs | Left behind | Access ends |
|---|---|---|---|---|---|---|
| 1 | 2004-2010 | On-premises accounting package | Open receivables, payables and balances only | Customers re-keyed, no mapping | A database backup on a retired server | Server sanitized in 2016 (IT lead) |
| 2 | 2010-2016 | Hosted ERP | Orders from 2012 in full; 2010-2011 as monthly summaries | Kept | Export files on a shared drive | No end set; the controller can open them |
| 3 | 2016-2022 | First cloud ERP | A five-year snapshot of transactions | Mapped | The vendor's tenant, read-only | Contract ends mid-2027 |
| 4 | 2022 onward | Second cloud ERP | Not applicable: current system | Not applicable | Not applicable | Not applicable |
Read across, the founding year promised about twenty-two years. The canvas shows order records that can be joined from 2012 onward (estimated), of which only the years from 2017 sit in the system the company uses every day. Orders from 2012 to 2016 survive in the first cloud ERP's read-only tenant and in export files. The years 2010 and 2011 exist only as monthly summaries, and nothing from before 2010 can be retrieved.
The summary line the CEO approved was short: "Order history joinable from about 2012 across three systems; current ERP holds 2017 onward; 2012-2016 in a read-only tenant until mid-2027 and in exports; 2010-2011 summaries only; pre-2010 records not retrievable."
Label estimates, constraints and the unknown
- Reported or estimated, era by era. Years reconstructed from memory are estimates; years read from a project file are reported. Say which.
- Dated constraints go in. The mid-2027 tenant end is a documented event and the reason to talk sooner rather than later. A sales quarter is not.
- Earlier owners are flagged, not resolved. If an era predates an acquisition, write that down; what it means for rights is a question for SourceX's review and the company's counsel.
- Unknowns get an owner. "Whether attachments crossed in 2016: unknown; the controller will check the project file." The company asks its own vendors if it needs to; the partner does not.
Nothing on the canvas is settled by pulling an export to see. A canvas with three honest unknowns is more useful in review than one with no gaps and no basis.
The canvas goes to the CEO first, then to SourceX
The CEO approves the canvas, the relationship owner shares the summary line with it, and the company applies itself through the partner's link. The partner who submits the canvas needs a signed SourceX Channel Partner Agreement beforehand. Credit goes to the earliest valid introduction made inside the 90-day attribution window, is assigned when the company submits a verified application and is decided from dated records.
A first call between the company and SourceX can walk the canvas left to right:
- The company describes each era and what crossed at each gate.
- It names the eras it would consider offering and where each one sits now.
- Access end dates are confirmed, with who holds each era.
- Open questions are assigned to people at the company, with dates.
- The company decides on a next step; nothing is sent before it does.
For the step after an introduction, guidance on what to do with old CRM data after a migration and on exporting data from a legacy system covers the company's own preservation choices. A worked accounting example is in moving historical data from QuickBooks to NetSuite.
Where a timeline canvas leads you astray
- Memory fills gaps. Eras before current staff arrived are reconstructed, and confident recollection is not a project record.
- "Everything" needs unpacking. Attachments, free-text notes and custom fields are worth naming one by one, because each can be left out of an otherwise complete move, and the context lost in SaaS migrations can matter as much as the rows.
- Acquisitions add eras. A company built from acquisitions may have several timelines, one per acquired business; system decommissioning after an acquisition is where those often end.
- Joinability is not the only test. Separate eras may still be useful on their own. The canvas records the split; it does not decide its consequences.
- Programs move. SourceX's current data needs differ in operating history and other criteria, so a company that misses today's program could still suit a later one.
How a platform team runs introductions across its whole portfolio is set out in the venture platform overview.
Next step
Refer a company. No cap on rewards. Default reward on a company's first qualifying transaction: the greater of $50,000 plus 5% and 15% of Actual Supplier Payout, with no cap; further qualifying transactions first agreed within 12 months of the first Closing pay 15%. Eligibility and payment conditions apply, including the USD 100,000 Minimum Transaction, and a signed Schedule A or addendum can differ. A lead, meeting or signature alone does not trigger payment.
If the canvas shows years of records the company can still open, and its CEO is open to a conversation, get a partner account, compare the summary line with the company criteria and read the terms of the program before anything goes to the CEO.
Whether SourceX itself merits a closer look as an investment is a question for a separate track from this canvas; investor materials can be requested through the SourceX investor overview.
This article is for information only and is not an offer to sell or a solicitation of an offer to buy securities. Any offering by SourceX is made only to verified accredited investors through its offering documents.